Schools are readily adopting prevention programs. While program alternatives are plentiful, schools face a serious dilemma because they lack the means to identify effective programs. The objectives of this project are (1) to statistically model the effectiveness of existing alternative prevention programs, (2) to develop an innovative intervention that uses information about program effectiveness to guide schools in adopting and implementing effective substance abuse prevention programs, and (3) to test the effects the intervention has on changes in program choice and effectiveness. This project will consist of the following activities. First, data will be gathered about the nature of existing programs being delivered to students. These data will be combined with student data on mediators and behavioral outcomes to identify strategies that hold promise for preventing substance use onset. From these data, a program will be created and tested that will use these results to guide the school prevention planning process. In this program, changing promising mediators will be promoted as being essential to effective prevention programming. Twelve schools will be assigned to receive either this feedback program or a traditional program that will provide teachers and principals with information about the biology of substance abuse. Schools will be stratified on key demographic features and randomly assigned to training condition. School decision makers will participate in assigned programs for three successive years. The effects of training will be assessed through monitoring the selection and implementation of programs and through evaluating program effectiveness in changing the onset of substance use among students. It is hypothesized that the feedback program will significantly improve program selection, implementation, and effectiveness. We believe this approach to prevention will serve as an archetype for the selection and implementation of effective school-based programs throughout the nation.